The discovery of the immunostimulatory activity of bacterial DNA and the base sequence: A specific base sequence (CpG DNA) containing an unmethylated cytocine/guanine dinucleotide (5′-CpG-3′) that occurs at high incidence in bacterial DNA is recognized by Toll-like receptor 9 (TLR9) and thus activates the immune system of a mammal to induce the Th1 immune reaction. It was reported by Tokunaga/Yamamoto et al. (Tokunaga T. et al., J. Natl. Cancer Inst. 72: 955-62, 1984) that the DNA fraction of BCG induces the production of type I interferon (IFN) and the subsequent activation of NK cells induced thereby and thus exhibit anti-cancer effect. In a series of studies, the investigators identified a palindrome type CpG DNA that is frequently found in bacterial DNA as the active sequence (Yamamoto S. et al., J. Immunol. 148: 4072-6, 1992). Several groups have demonstrated the immunostimulatory activity on mouse and human B cells by Escherichia coli DNA or non-antisense DNA, and Krieg et al. (Krieg A M. et al., Nature 374: 546-9, 1995) reported 5′-PuPuCpGPyPy-3′ that has unmethylated CpG and that is flanked by specific bases as a mouse B cell-activating motif. Other unique sequences of CpG DNA that exhibit an immunostimulatory activity have been presented and, at present, they are roughly grouped into the IFN induction type, the B cell activation type, and the mixed type (Verthelyi D. et al., Trends Immunol. 24: 519-22, 2003).
On the base sequence that activates mouse NK cells: In order to identify active base sequences, the Tokunaga/Yamamoto group selected at random and synthesized base sequences of a certain length from cDNAs that encode the BCG protein. After investigating 30-chain length bases 5′-ACCGATNNNNNNGCCGGTGACGGCACCACG-3′ (SEQ. ID. No. 1) (N is a complementary base pair), they thought it important that the N portions include CpG and that three consecutive bases on one side are followed by the complementary bases forming a palindrome structure (Yamamoto S. et al., J. Immunol. 148: 4072-6, 1992).
Krieg et al. discussed whether, as the mouse B cell activating motif PuPuCpGPyPy (specifically, GACGTT is potent) also has an activity of enhancing the NK activity, the important sequence for the activation of NK cells is the unmethylated CpG which is followed by a specific base, but that the motif does not have to take a palindrome sequence (Krieg A M. et al., Nature 374: 546-9, 1995; Ballas Z K. et al., J. Immunol. 157: 1840-5, 1996). Boggs et al. reported that the CpG dinucleotide alone, though required, cannot activate NK cells, and specific bases and background sequences surrounding CpG and their modifications such as thiolation or methylation define the NK cell activation by CpG DNA (Boggs R L. et al., Antisense & Nucleic Acid Drug Development 7: 461-71, 1997). For example, the thiolation of CpG DNA leads to an attenuated activity of activating NK cells.
The methylation of the active motif of CpG leads to decreased NK activity. However, depending on the base sequence of the motif, activity may be retained even after methylation of CpG. In this case, the methylation of all cytosines in CpG DNA leads to total annihilation of the activity. As the motifs of CpG DNAs so far reported to have an activity of enhancing the NK activity are not limited to palindrome or PuPuCpGPyPy and are composed of 6-chain length centering on the unmethylated CpG surrounded by specific base sequences, it is believed, the enhancement of the NK activity by CpG DNA is induced by a higher structure constructed by the entire base sequence comprising the active motif, the background sequence and modifications.
Reaction of mouse NK cells to CpG DNA requires activation: Yamamoto et al. have demonstrated that the enhancement of the NK activity by CpG DNA is mediated by type I IFN that was produced by cells other than the NK cells (Yamamoto S. et al., J. Immunol. 148: 4072-6, 1992). Klinman et al. report that the IFN-γ-producing cells in mouse spleen cells induced by CpG DNA are NK cells and the IFN-γ production is inhibited by IL-12 antibody (Klinman D M. et al., Proc. Natl. Acad. Sci. U.S.A. 93: 2879-83, 1996). Halpern et al. demonstrated that bacterial DNA or CpG DNA stimulate monocytes/macrophages to induce the production of IL-12 or TNF-α, with a result that IFN-γ is produced by non-adhering cells (Halpern M D. et al., Cellular Immunol. 167: 72-8, 1996).
Ballas et al. have demonstrated that NK cells respond to CpG DNA in the presence of IL-12, IFN-α/β and TNF-α, and thus the NK activity becomes enhanced. Chace et al. also report that the activation of NK cells is macrophage-dependent and that NK cells, when activated by IL-12, acquire reactivity to bacterial DNA and IFN-γ production is amplified (Balla Z K. et al., J. Immunol. 157: 1840-5, 1996). These results suggest that inactivated mouse NK cells cannot react to CpG DNA but are activated by stimulation with cytokines derived from CpG DNA-stimulated monocytes/macrophages, and thus acquires CpG DNA reactivity.
Activation of human NK cells by CpG DNA: The effect of CpG DNA is conspicuous in the mouse immune system but the reactivity is generally low in the human immune system. However, the anti-cancer activity of BCG DNA mediated by host immunity had been reported in the 1980s, and the NK activity is enhanced both in vivo and in vitro. On the other hand, human peripheral blood mononuclear cells (PBMC) produce IFN-α, IL-12 and IL-18 by stimulation with CpG DNA, and the production of IFN-γ is induced by IL-12 stimulation (Bohle B. et al., Eur. J. Immunol. 29: 2344-53, 1999). These results suggest that, in humans as well as in mice, dendritic cells and monocytes/macrophages are involved in the activation of NK cells by CpG DNA.
Iho et al. speculated that the action mechanism and the activity sequence of CpG DNA in humans are different from those in mice because stimulation of human mononuclear cells by BCG DNA does not significantly induce the production of IL-12. In fact, among the PuPuCpGPyPy type CpG DNAs that are reported to be active in mice, #1643 (gagaacgctcgaccttcgat) (SEQ. ID. No. 2) that activates B cells, #1618 (tccatgacgttcctgatgct) (SEQ. ID. No. 3) that induces IFN-γ, an antisense DNA #1758 (tctcccagcgtgcgccat) (SEQ. ID. No. 4) that activates NK cells, and #2105 (ttgcttccatcttcctcgtc) (SEQ. ID. No. 5) that activates human B cells were investigated for activation of NK cells with IFN-γ production as an index, but none of the sequences activated the purified human NK (CD56+) cells.
Thus, among cDNAs that encodes the BCG protein, ten 30-chain length DNAs (Iho S. et al., J. Immunol. 163: 3642-52, 1999), accgatNNNNNNgccggtgacggcaccacg (SEQ. ID. No. 6), that contain specific six-chain length CpG palindrome (NNNNNN) were investigated, and seven of them were found to have the IFN-γ-inducing activity. Then, mouse B cell-activating motifs aacgct and aacgtc that are not effective for the activation of human NK cells were inserted into the background sequence of CpG DNA that was activated with human NK cells to synthesize accgataacgctgccggtgacggcaccacg (SEQ. ID. No. 7) and accgataacgtcgccggtgacggcaccacg (SEQ. ID. No. 8) (underlined sequences were inserted), and they were investigated for activity, but no activity was recognized. This indicated that for CpG DNA to be recognized by NK cells the activity motif of CpG DNA must assume a palindrome structure.
However, as even for the same sequence there was variation in the amount of IFN-γ produced between donors, and there were individual differences for the optimum sequence, 12-chain G was added (g12CGA) on both sides of the underlined palindrome portion CGATCG for the weaker accgatcgatcggccggtgacggcaccacg (SEQ. ID. No. 9) and they were investigated. The reason for adding G is that poly G has a high affinity for cells and stabilizes the higher structure of DNA. As expected, G addition enhanced the IFN-γ-inducing activity of CGATCG. When sequences having further higher activity were investigated, higher activity was noted in a full-length 30-chain g10GACGA (GGGGGGGGGGGACGATCGTCGGGGGGGGGG (SEQ. ID. No. 10)) in which the palindrome portion of g12CGA was repeat-extended to 10 bases and 10-chain G was added to each side chain.
G10GACGA is a sequence that has the most potent activity among the human NK cell-activating sequences reported in 1999. As described below, though NK cells when activated exhibit a high reactivity to CpG DNA, the induction of IFN-γ by G10GACGA does not disappear after neutralization of non-NK cell-derived cytokines with anti-IFN-α antibody etc. and, therefore, it is thought that human NK cells are equipped with an ability to react to CpG DNA. IFN-γ produced by G10GACGA enhances the NK activity through an autocrine reaction and induces the expression of CD69 and HLA-ABC.
CpG DNA reactivity of activated human NK cells: Activated human NK cells have a higher reactivity to CpG DNA as compared to unactivated NK cells (Iho S. et al., J. Immunol. 163: 3642-52, 1999). This is in agreement with the fact that the expression of TLR9, a CpG DNA receptor, is weak in non-stimulated NK cells (Krug A. et al., J. Immunol. 31: 2154-63, 2001). However, there are no reports that TLR9 is induced by the activation of NK cells, and the mechanism in which CpG DNA reactivity is enhanced by activation is unknown. Furthermore, as NK cells activated by IL-2 also react to the non-palindrome type CpG DNA (Iho S. et al., J. Immunol. 163: 3642-52, 1999), the sequence selectivity of CpG DNA must be investigated.
The difference in CpG DNA reactivity between human NK cells and mouse NK cells: Human NK cells, even if unactivated, react to CpG DNA, but mouse NK cells do not react to CpG DNA unless activated. The reason for this difference in reactivity is not known. Ballas et al. (Ballas Z K. et al., J. Immunol. 157: 1840-5, 1996) investigated a CpG palindrome having a G-repeated sequence on both of 5′-end and 3′-end, but it could not activate mouse NK cells nor human NK cells. #2216 of Krug et al. (Krug A. et al., J. Immunol. 31: 2154-63, 2001) has GACGATCGTC but does not activate human NK cells. The difference between the present g10GACGA and the CpG DNA of Ballas/Krug et al. is the number of poly G added and modification thereof. The former is an unmodified type, while part of the latter has been thiolated.
The thiolation of DNA enhances resistance to DNase, but lowers the immunostimulatory activity because interaction with DNA-binding protein is weakened. Thus, it is likely that thiol modification may cause low CpG reactivity of NK cells. On the other hand, some unmodified CpG palindrome having poly G in the background sequence like ggggggggggggaacgttgggggggggggg (SEQ. ID. No. 11) have no activity (Iho S. et al., J. Immunol. 163: 3642-52, 1999). The sequence and length of the palindrome bases are considered important elements. It was later reported that the reactivity of human cells to the mouse activation motif is low, and reactivity to CpG DNA in primates was found to be different between humans, chimpanzees and monkeys, and therefore it is generally accepted that there is a species difference in CpG DNA reactivity (Hartmann G. et al., J. Immunol. 164: 1617-24, 2000). It is also becoming clearer that the CpG DNA sequence has cell selectivity (Verthelyi D. et al., Trends Immunol. 24: 519-22, 2003).
Intracellular incorporation of CpG DNA and recognition by TLR9: A recent study has revealed that CpG DNA is a ligand for TLR9 (Hemmi H. et al., Nature 408: 740-5, 2000). In an examination with a confocal microscope, it was shown that the binding of CpG DNA to the cell membrane and its incorporation into the cell are not sequence-specific and CpG DNA is localized together with TLR9 in the endosome. Thus, it is believed that CpG DNA is incorporated into the cell by endocytosis and it is recognized by TLR9 in the endosome. In order for CpG DNA to exhibit its biological activity, it must be modified in some way in the endosome. In the process of CpG DNAs being incorporated into the cell and recognized by TLR9 and of TLR9 signal transduction, a plurality of molecules need to work in concerted actions and research is on going.
Signal transduction of CpG DNA: While the incorporation of CpG DNA into the cell is effected in a CpG-nonspecific manner, the process from recognition by TLR9 to the expression of biological activity is CpG-specific. Cells that strongly express TLR9 in human peripheral blood are mainly B cells and plasmacytoid dendritic cells (PDC) (Hornung V. et al., J. Immunol. 168: 4531-7, 2002). In CpG DNA-stimulated B cells, the activation of p38 and JNK occurs very early and the ability of a transcription factor AP-1 to bind to DNA increases and the transcription of related genes are enhanced Hartmann G. et al., J. Immunol. 164: 944-52, 2000).
In PDC, CpG DNA is incorporated by endocytosis and then is recognized by TLR9 thereby to activate p38 MAPK. Subsequently, STAT1 is phosphorylated to form ISGF3 together with STAT2 and IRF-9. This leads to the enhanced transcription of the IRF-7 gene, and the IRF-7 thus produced induces the transcription of the IFN-α gene to produce IFN-α. Furthermore, extracellularly secreted IFN-α is fed back to stimulate the JAK-STAT pathway and thus a large quantity of IFN-α is produced (Takauji R. et al., J. Leukoc. Biol. 72: 1011-1019, 2002).
Significance of the poly G-added palindrome CpG DNA: CpG DNA is recognized by target cells centering on at least six-chain bases containing the CpG dinucleotide, and the activity greatly varies with slight differences in not only the core sequence but the surrounding bases (background bases). In order to directly induce PDC to produce IFN-α and to produce a large quantity of IFN-α by an autocrine reaction, it is important that the core sequence of CpG DNA takes a palindrome structure. Then, the length and the position and the types of bases on both sides are mentioned as factors that affect activity. In fact, the consecutive addition of G having a high affinity for the cell membrane to the core sequence “GACGATCGTC” (SEQ. ID. No. 12) induces a high activity.
On the other hand, poly G per se inhibits the IFN-γ production in mouse spleen cells (Halpern M D. et al., Immunopharmacology 29: 47-52, 1995), the addition of G has a risk of affecting the loss or inhibition of activity. This, in a different perspective, means a possibility that by changing the palindrome sequence and the mode of adding G, the CpG DNA activity may be regulated and CpG DNA that selectively induce specific cytokines may be developed. In fact, the poly G-added palindrome CpG DNA strongly induces the production of IFN-α or IFN-γ in PBMC but does not induce that of IL-12 or IL-6. In a later study this sequence was termed the D or A type oligo, and G10GACGA reported by Kuramoto/Iho et al. in 1992 (Kuramoto E. et al., Jpn. J. Cancer Res. 83: 1128-31, 1992) and 1999 is included in this type.
A DNA sequence composed of consecutive guanines is called a poly G sequence or a G-quartet, and enhances the incorporation of CpG DNA into the cell. Thus, it is believed that in the poly G-added palindrome CpG DNA a double stranded overlapping region is formed by the palindrome base sequences, and the poly G added to the end thereof enhances resistance to nuclease digestion and stabilize the higher structure of DNA so that activity may be efficiently developed. In fact, the CpG palindrome DNA in which poly G was introduced has a high activity of inducing IFN-α or CXCL10. Thus, it is thought that a Th1 immune reaction is efficiently induced and thus usefulness for application to the treatment of cancer, allergies and infections may be expected. In animal experiments, a thiol-modified CpG DNA is usually used, which is reported to bring about fatal side effects. From these facts, it is important to develop an unmodified poly G-added palindrome CpG DNA that is safe and has a highly effective Th1 immunostimulatory activity.
Patent document 1: Kohyo (National Publication of Translated Version) No. (A) 2002-510644
Patent document 2: Kohyo (National Publication of Translated Version) No. (A) 2002-517156
Non-patent document 1: Tokunaga T. et al., J. Natl. Cancer Inst. 72: 955-62, 1984
Non-patent document 2: Yamamoto S. et al., J. Immunol. 148: 4072-6, 1992
Non-patent document 3: Krieg A M. et al., Nature 374: 546-9, 1995
Non-patent document 4: Verthelyi D. et al., Trends Immunol. 24: 519-22, 2003
Non-patent document 5: Ballas Z K. et al., J. Immunol. 157: 1840-5, 1996
Non-patent document 6: Boggs R L. et al., Antisense & Nucleic Acid Drug Development 7: 461-71, 1997
Non-patent document-7: Klinmann D M. et al., Proc. Natl. Acad. Sci. U.S.A. 93: 2879-83, 1996
Non-patent document 8: Halpern M D. et al., Cellular Immunol. 167: 72-8, 1996
Non-patent document 9: Bohle B. et al., Eur. J. Immunol. 29: 2344-53, 1999
Non-patent document 10: Iho S. et al., J. Immunol. 163: 3642-52, 1999
Non-patent document 11: Krug A. et al., J. Immunol. 31: 2154-63, 2001
Non-patent document 12: Hornung V. et al., J. Immunol. 168: 4531-7, 2002
Non-patent document 13: Hartmann G. et al., J. Immunol. 164: 1617-24, 2000
Non-patent document 14: Hemmi H. et al., Nature 408: 740-5, 2000
Non-patent document 15: Hartmann G. et al., J. Immunol. 164: 944-52, 2000
Non-patent document 16: Takauji R. et al., J. Leukoc. Biol. 72: 1011-1019, 2002
Non-patent document 17: Halpern M D. et al., Immunopharmacology 29: 47-52, 1995
Non-patent document 18: Kuramoto E. et al., Jpn. J. Cancer Res. 83: 1128-31, 1992